Any of the 7 dBm LO mixers should work, but avoid the ones designed for VHF/UHF work.

Use the IF port (which is good to DC) for the frequency out--the IF and RF ports are usually limited BW due to the ferrite transformers used on those ports. I'd add a -6dB attenuator on the IF port, just to keep it reasonably terminated--the IF port is the most critical for termination impedances.


Pete

I had another idea. Get a softrock single band receiver kit (fivedash.com) for 21 dollars. Then change the crystal to a round number frequency near the kit center frequency. Instant converter. The outputs are good to about 200Khz I think and you could adjust a few caps to go higher. I have the ensemble II LF sitting on the shelf since I got my AFEDRI. I think I will try that. It has the drawback of needing 12 volts and 5 volts but you don't need the extra cable connection and you don't need to worry about blowing up the mixer. The Ensemble II can be programmed to power up at any frequency within its tuning range so you don't need a computer connection after programming it once. But it costs more (about 70 bucks I think). The kits use some surface mount parts so they can be a bit of a pain to build. I did it though and I have a pretty good tremor.

I tried my softrock and it works fine. I can get 0dbm out no problem. At 0dbm out, 2nd harmonic was 70db down and 3rd was 65 db down. It is about -18dbm in for zero dbm out. At +7dbm out 3rd harmonic distortion is still OK at 50dbm down but at +8dbm out, distortion is -11dbm. Not bad. This was at 50Khz out. At 125Khz it is still good but it takes -11dbm to get 0dbm out. At 200Khz it takes -5dbm in for 0dbm out.

Tony

Another way to accomplish the goal, albeit somewhat more elaborate.

So how is the DBM in any danger here? You probably have a bit more than enough LO signal from the generator/counter timebase output to drive the mixer, and the output of the mixer IF port will be about 6 db down from the generator's indicated RF output level, and NO power supplies, software, or computer are needed.

At one time there were a few vendors selling NE602/NE612 mixer boards. Single chip balanced IC mixer, with an internal oscillator section. With one of those boards, and a one or two MHz crystal, and one could make a pretty cheap down converter for under twenty bucks.

I used my last 602 a few weeks ago, and bought fifty more from HK (#330954746888) earlier this month, great little IC, and not expensive. But again you need a supply, or at least a battery to use with it.

Pete,
Maybe the DBM is not in danger. I was under the impression that overdriving them could damage the diodes. I will probably forget I have it connected and flip the 3 volt output switch.

Tony

Check out HP App Note AN171-2: Extending the 8640B to DC.

Well, I built the MCL Mixer circuit using an SBL-1 mixer from eBay. I can't find the data sheet for this variant (SBL-1-19) but the seller says it is .1 to 500 Mhz.

It works fine but there is about 30db of loss through the whole shebang. I built it with a 6db 50 ohm pad on the IF output and measure it on the spectrum analyzer that has a 50 ohm input impedance. The second and third harmonic (of the output frequency) are more than 40 db down when the RF input is -20dbm. At -10dbm input the third harmonic is 30db down. Some higher harmonics (around 2Mhz I think) are actually much stronger but should be easier to filter if necessary.

If you need more than -30dbm out, then you need to add an amplifier and the distortion is getting significant at this level.

Note that this is on a Boonton 103D using the 1 Mhz reference output for the LO.

Tony

30 dB is excessive. 8 dB for the mixer,and another 6 for your attenuator. It shouldn't be more than 15 dB loss.

Are you using the IF port for the output--that is the only port that is good to DC.

Pete

Yes, I am using the IF port for the output. At the output I have a PI network 6db pad consisting of two 150 ohms in the legs and 39 ohms across the top of the PI. LO and RF inputs as well as the output to the spectrum analyzer are coupled through 1uf ceramic MLCs. I tested it down to 10Khz. What if the LO input is weak? Remember I said that the reference output dropped to -7dbm with a 75 ohm load.

Tony

Tony C said;

Have you measured it with the DBM as the load?

I was curious which mixer you have, but I see you mentioned it in your last post. Here is the mini-circuits site with a list of parameters for the various types.

This model (with the -19 variation) is not listed anywhere on the MCL site.

Yes I measured the LO with the DBM hooked up right after my previous post. It measures about -9.5dbm. I'm pretty sure that is the reason for the low conversion gain. I could add a buffer but then I need power.

Tony

It will go into gain compression at a much lower signal input level with a low LO injection. Those are designed to be hit pretty hard on the LO. I suspect conversion losses might increase if the LO is too low.

Pete

Just for grins I checked the output levels of several generators and my counter, into a 50 ohm input on my scope. With the exception of the 3336 all are fairly nice sin waves.

Model.........Freq..........RMS level

HP3586.....10Mc...........500mV

HP5245M...5Mc.............1V

HP8660.....10Mc............770mV

HP8657.....10Mc............420mV

HP3336.....1Mc.............750mV P-P Sq wave

Also hooked up a MCL SRA-1 with the 5Mc LO from the 5245M counter optimized for best loss/spur levels, and a 5.5Mc signal at 0dbm. Conversion loss about 9db and close in spurs at -40db or so.

First spike is "0", second is IF out 500kc away, 5Mc lo is about mid screen at about +4db.